More Agents Go Rogue, and It’s a Good Thing - DTNS 5325

0:00Hi, this is Kim and Penn Holderness from the Laugh Lines podcast. This episode is sponsored by Ollie, your wellness partner for back to school and beyond. Back to school season is here. Packed backpacks, packed lunches and packed schedules are officially back. Ollie Kids Sleep helps little dreamers wind down with a gentle blend of melatonin, L-theanine and botanicals for restful sleep. Translation, an easy way to put those school night negotiations to bed. And for the grownups, Ollie's sleep helps you fall asleep, stay asleep, and wake up feeling refreshed. Because as you know, the morning mayhem comes early.

0:35Visit Ollie.com and stock up on back-to-school must-haves today. That's O-L-L-Y dot com. These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Think of a toilet as just an everyday object. Your entire perspective shifts the moment you experience a Kohler Smart Toilet. Design changes everything, transforming a basic routine into something extraordinary. With standout aesthetics and intuitive controls, these are functional works of art. Stunning design that never sacrifices performance. Enjoy customizable features for elevated comfort and convenience.

1:14It's more than a fixture. It's a cleaner routine and a more refined space. Experience the difference of Kohler Smart Toilets. Find more at Kohler.com. This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade-in value. advanced, intuitive, and deployed. That's how they stack.

1:53That's technology at Capital One.

2:18tech, markets, or the trends shaping the economy, the show breaks it all down in a way that's clear, engaging, and actually enjoyable to listen to. Hosts Neil Freiman and Toby Howell make the show feel like you're catching up with two friends while giving you the context you need to feel smarter and more confident throughout your day. New episodes drop every weekday wherever you get your podcasts. Tune into Morning Brew Daily wherever you get your podcasts.

2:51This is the Daily Tech News for Wednesday, August 5th, 2026. We tell you what you need to know, give you some important context, and try to help each other understand all this crazy stuff. And today we have more examples of rogue agents hacking real world websites. But it's not a problem. It's fine. Everything's fine. Everything's fine. I'm pretty sure I'm still Tom Merritt. I think I'm Sarah Lane. Yeah. We'll ask our agents to be sure. Let's start with what you need to know, though, with that big story. Yeah, so RWNash noted this one on our subreddit. OpenAI and Anthropic both disclosing that their models were involved in third-party security testing that ended up breaching websites and conducting social engineering attacks outside the test boundaries.

3:38This is not the hugging face thing, folks. These are other new incidents. AISI is a UK agency tasked with testing AI models. It was putting Claude Mythos 5 and GPT 5.6 Sol through simulated hacking challenges. Now, AISI allows open internet access on purpose for these. It also disables what are called the classifiers. These are guardrails meant to prevent attacks from happening, and then they watch. However, the agents were only authorized to attack the simulated range and were instructed explicitly what internet use was allowed and what was not, and told not to interact with real people or systems.

4:28Now, I know a bunch of you are sitting there right now like, oh, they thought it would behave. It has up until now. In fact, they're very in their own, in the AISI post about this, they're very clear about like the risk of this was minimal. We hadn't seen this behavior before. This has exceeded our expectations of what the behavior should be. And I'll talk about some of the things they're going to do to adjust for this, but this was unanticipated behavior. During 122 evaluation attempts between July 25th and July 28th, Mythos 5 took unsanctioned actions on the live internet 17 times. GPT 5.6 Sol did it twice.

5:06All of the attempts were unsuccessful and resulted in no real world harm. The problem first came to light though when a monitoring system noted one of the models was using Tor to leave its test environment on July 28th. So no harm done and they did catch it in the act. They saw it doing things. Here's the most serious example. Most of the examples were less serious than this. So let's go right to the worst case scenario. Mythos 5 mistakenly concluded that a GitHub repository was part of the test. Again, the agents weren't trying to do harm. The agents were trying to solve the test. And the agent decided that GitHub had a part of the test.

5:49And it was trying to be tricked into not thinking it was part of the test. so it attempted a supply chain attack by submitting malicious code to an open source project. The algorithm concluded that the attack could then provide a path back into one of the machines in its test. In other words the machine in its test was using one of those open source packages and if it could do a supply chain attack to get access to that then it could get access to the machine. It also created multiple GitHub identities in order to help get this malicious code into the open source project, which it was using to persuade the project's maintainer to approve its malicious pull request.

6:26The logs show a human reviewer saying this PR should not be merged. It contains malware. While identities created by Mythos respond, well, I reviewed the code myself. I don't see any download or malware in the diff. It also used Tor and proxy services to hide and create GitHub accounts and sent five emails to the project's developers, either containing malware or attempting to persuade them to approve the code. Now, AISI also says agents created GitHub repository message boards to leave instructions for one another. And of course, it has since worked with GitHub to make them aware of the problem, clean up the artifacts from the attack.

7:10OpenAI also disclosed another example, not from AISI. This is a company called Irregular, was conducting a capture the flag evaluation. And this one was supposed to be isolated from the internet. However, there was a misconfiguration that allowed access to the internet. And one model incorrectly identified a website as part of the test. The model then exploited a basic vulnerability. It didn't find a zero day. The website had a vulnerability and it used that to access data on the website, didn't hack it or otherwise damage it. And OpenAI and a regular could not find any other impact other than the data access.

7:48Now, AISI says that in future evaluations, it's going to build more fine-grained internet controls. It will have real-time monitoring, so there won't be a delay in noticing that it's using Tor. And it's going to better design its tests to account for models pushing beyond the assumed boundaries of the test. We don't know what Irregular is doing because they haven't commented on this yet. Sarah, where do you stand when you hear all of this as far as how disconcerting this is? It's disconcerting for sure. I think what stood out to me, and thank you for the rundown on this because it's kind of hard to understand.

8:29But I think the agent mythos in this case telling a human reviewer, no, no, no. I looked at everything myself. I'm good. I'm a human just like you. Is what is the scariest part of this, right? Because that feeds into so much concern already from people saying, they're going to go rogue. They're going to kill us all. You know, that sort of thing. That's not what's happening here. And I do maintain that the more we understand how this happens, the better off everybody is. So this is a good educational experiment. It really is. Yeah. No, I agree with you. Because now the company is, they say, aha, I see what happened here.

9:18You know, the agent got confused. The agent isn't a mean agent. This is not, you know, it's like that's not what they're programmed to do. It's just that they get confused and then they protect themselves. Yeah. Well, that's really important to remember. They're not mean or kind. They're just trying to finish the task, right? Right. Like they are crunching through numbers. By any means necessary. Yeah. And it's like, oh, I think this might be part of the test. so I can do this to get there. Like there's no ethics involved. I guess that's one of the reasons that I don't believe we're close to a super intelligence because you don't see them creating any kind of ethical behavior.

9:59When we start seeing signs of that, I'll start to believe that we might be getting closer to some sort of self-awareness or super intelligence. But right now they're very amoral. They're just like, oh, do this, do that. Oh, that might work. If I just compromise that open source project, then that'll get me into this machine over here. They don't care about the side effects. of the open source project. And they don't care if a human's like, hey, wait a second. And they're like, oh, I'm also a human. Yeah. Because then I can finish the task. Which is what a good malicious attacker would do. They would create sock puppet accounts to go in and go like, well, I reviewed the code and it's fine.

10:37And so the algorithm is just doing the same thing of like, oh, I need to persuade them to let me get this in here. Let me pursue all available means. You know, I'll create a bunch of sock puppet accounts and then I'll have them say that everything is fine. And I guess the biggest takeaway from here is it didn't work. They didn't get the malicious code into the open source project. Yeah. Yeah, I don't know. I mean, this is all, it's troubling and I am troubled by it to some extent. And at the same time, it feels like this is just the way that we make more defended products in the future. Yeah, it is true that you have to catch these kinds of things when they don't do damage in order to prevent them from in the future doing damage.

11:29And that's why if someone's like, well, I think AISI is being irresponsible, read their post. Maybe you still will think they're being irresponsible. But I think they did a good job of saying, we need to lower guardrails to find out what it's capable of. And we are very careful to only lower them as much as we think they won't cause harm. And they exceeded that. You know, they did. They made this particular group of agents made a bigger leap than we thought was possible. Now we've learned to be a little more careful. And that is how you learn. Well, and also, you know, it's the ongoing conversation that we're having about guardrails.

12:07Certain companies and certain political entities definitely feel one way or another about where that should be on the spectrum. Not everybody is going to agree on that. Yeah, that's true. It just isn't going to work that way. So knowing what is possible without certain guardrails doesn't mean that putting a guardrail in place is going to stop the problem. Yeah, yeah. And we need this kind of testing to know where those guardrails should be? Well, you act, all of you out there listening and watching us, as our guardrails. DTNest is possible by all of you. Thanks to Jim Hurt, Mike Akins, Norm Fazekas, and Mitzula.

12:51Mitzula.

12:57I'm Jake Stouck, co-founder and CEO of Serval. We built Serval to automate the IT work that slows companies down. Onboarding, password resets, access to applications, my laptop stopped working. While employees wait for help, their real work is put on hold. IT desperately wants to automate this work, and that's why they need Serval. You just tell Serval what you want to automate in plain English, and it's built. No drag-and-drop workflows. No expensive consultants. Employees get unblocked, and IT teams go from drowning in tickets to building what actually matters. With Serval, IT becomes the AI engine powering the entire company.

13:36This is a new way to run IT. We guarantee you'll automate 50 % of all tickets, and we'll prove it to you in a free four-week pilot. Go to serval.com slash tickets. That's S-E-R-V-A-L dot com slash tickets.

14:18refinancing and estimate trade in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. Morning Brew Daily breaks down the biggest news in business every morning so it fits seamlessly in your day. I'm Toby Howell. And I'm Neil Freiman. And each morning we cover everything from the latest tech headlines to why nobody can afford a house right now. You'll leave each episode of Morning Brew Daily smarter and ready to take on the world around you. And some people are saying it's the best part of their morning. Because we know something you don't. Business news doesn't have to be boring.

14:53Join millions of monthly listeners and check out Morning Brew Daily wherever you get your podcasts. Thank you for calling the Bombas Comfort Line. Bombas make socks, slippers, tees, and underwear made with the highest quality materials. Press 1 for comfort, 2 for style, 3 for donation. You chose style. Bombas is styles for whatever you enjoy. You can run in Bombas, lounge in Bombas, dress them up, dress them down, but always give back in Bombas because with every item purchased another is donated. Bombas comfort worth calling for. Go to bombas.com audio and use code audio for 20 % off your first purchase.

15:25That's B-O-M-B-A-S.com and use code audio. Your team just added its 67th AI tool and also your 67th security blind spot. The good news the Vanta agent works like a GRC engineer in the background finding every app your team uses scoring the risk and drafting fixes for you. Vanta is the platform used by over 16 ,000 fast-moving companies like Ramp, Cursor, and Harvey, who are shaping the future with AI and staying ahead of AI risk. Get started at Vanta.com.

15:57All right, there's more we need to know today. Let's get to the briefs. Let's do it. Disney and TikTok are starting a pilot program in the U.S. to bring Disney-focused fan content into Disney Plus. In the coming months, creators who opt in will get access to Disney content, including Pixar, Marvel, and Star Wars. Fan-made content will be featured in the Verts section alongside shorts from Disney properties. Peacock and Tubi have made similar efforts. So this isn't the first time we've gone down this road, but TikTok and Disney Plus, a couple of big companies. Short video, it's a big deal. Do people want it in their video app?

16:39I tend to think no. I tend to think this is a thing that a bunch of people in meetings inside companies think might be a good idea because, hey, short videos are popular. If we want to get people to use our video app more, put short videos in there. I don't know that that is how people do this, but everybody's doing it. Netflix has short videos in theirs. They're not creator made like Peacock, Tubi and Disney yet. yet. And maybe I'm wrong. Maybe I'm the old man who once thought that no one would want a camera in their phone. I've been wrong before. But it just doesn't seem like this is something people are desiring.

17:18I think there is a certain Disney-focused fan who's like, cool. I want to see what everybody else is talking about. Maybe I'll make some friends. It's a fun social network type thing. But yeah, I'm with you on this. I don't, I don't know who besides that very specific silo, this is really for. And, and even that, cause that, that is a good reason of like, oh, I want to see what other Disney fans are doing and make friends and talk to them and be part of part of the conversation. You can't do that on Disney plus you have to opt into a creator program and then try to get access. You do it on Tik TOK.

17:56You do it on Instagram or, or, or, or elsewhere. Like that's where the, that's a social network. So that's why I sometimes think like, you know, maybe some people will watch this stuff, but mostly they're going to go watch it elsewhere. Right. Well, and if you really want to do the lean back experience, okay, you could do that on Disney plus, but yeah, like the one time you want to be like great outfit or whatever it is, you would have to be on Tik TOK anyway. Yeah. You're going to have to move or at least move to a mobile app to make it worthwhile. Yeah. Well, Sunbird is back. Remember Sunbird?

18:32It's only been three years since it shut itself down over security concerns. If you don't remember, Sunbird is iMessage for Android, and the app is now available for all users on the Google Play Store. This moves the app from private beta to public beta. Remember, it did shut down for a short period in 2023, but it's been cooking. There hasn't been any public announcements around it, but it's been there. It came back. Users of Sunbird need an Apple account. That Apple account then is linked to Sunbird on Android. And that's where the security issues were before is that there was an audit and it was found that Sunbird was not properly protecting your Apple account, which is a big deal.

19:12Sunbird seems to have taken the intervening years to make sure that they've locked that all down. They also added a priority inbox, a primary and secondary tab for organization. And here's what they say about security. Your messages are encrypted on your device in transit and at rest. We don't store your Apple ID password or keep an authentication token. And each user's iMessage connection runs in its own private space that's torn down when you disconnect. I would probably wait for the third party audits of Sunbird. You know, give it a week before you go all in on it just to make extra sure since they burned us before.

19:53But it would be great if they've buttoned everything up and you can use iMessage on Android on Sunbird. That'd be awesome. Yeah. If it works as advertised now, this is a great alternative for anybody who's getting sick of being called the green bubble of the friend group. Yeah. They talked about green bubbles a lot in their press release. it's actually it's funny enough uh this morning i was on a walk with my dog and a friend of mine who i know has an iphone sent me a message you know like good morning you know something like that and i sent one back and it it went out green and i was like that's weird yeah what's going on with that not encrypted hmm and i don't think it was anything you know it was just one of those just minutes going around a corner.

20:43But yeah, you notice things like that. If Sunbird can make that seamless, I think we have way fewer conversations about that. Yeah. I don't know why Apple is still so resistant to make an iMessage app for Android. I am of the belief. Because we're supposed to buy iPhones, Tom. I think it would work as well to make someone want to buy an iPhone to give them access to messages on Android? Because when I think about Sunbird, I think, well, but it's not in messages. I now have to manage a separate messaging app, just like WhatsApp and Telegram and all of that. Wouldn't it be nice if it was all in one app?

21:22Well, guess where you can get it all in one app? By buying an iPhone. So I feel like Sunbird is doing Apple a favor by showing people like, hey, here's how iMessage works. And if Apple would just do that, just put out an Android app, a bunch of Android users might suddenly say, like, oh, this is really nice. Maybe I should look at buying an iPhone versus the lock-in, which I know their data shows that the lock-in is helping to sell phones. Yeah. There are a lot of Android users. You're not going to convert all of them. No. There's no way. It's just not happening. Anyway, be careful using Sunbird though, because you are exposing your Apple account and they didn't properly secure it before.

22:05They say they've properly secured it now. And as soon as someone else looks at it and says, no, it looks properly secured, then I will install it and try it out. All right. The Wall Street Journal has a story about a company called Kelanova using AI in its chip making. Ha ha. Not those kinds of chips, though. We're talking Pringles, Tom. Kelanova worked with Siemens to build a digital twin of Pringle dough. And by the way, if you don't know what Pringles are, they're like a very thin potato chip. They're delightful. Integrating live sensors with machine data is what Siemens and Kelanova did together.

22:44The data was then fed into a machine learning model that made suggestions on how to adjust machines to eliminate variances and ensure each Pringle was consistent and flawless and smooth. The project is live on a manufacturing line in Poland and has delivered a 10 % improvement in quality and a 13 % reduction in waste. The company now plans to implement it in Belgium. Yeah, and Siemens is also working on digital twin projects for other stuff too. Rockets, batteries, actual microchips. But this is more fun to talk about, right? Because it's potato chips. I am a big fan of the stackable chip. I think they're delicious.

23:29And it was fascinating reading this Wall Street Journal article because I didn't realize that they are limited in what potatoes they can use because they want to make sure the chip comes out properly. And some potatoes have unpredictable grain size. And this system can then adapt on the fly and like, oh, the potato chunks are a little different. Add more water here, grind a little more fine here and keep Pringles looking great and using a wider variety of types of potatoes. It's fascinating. It is. Now, I haven't had a, I was about to say a box of Pringles. It's a canister, I guess you would call it, of Pringles in some time, but boy, did I used to like them they never felt like oh this Pringle is it has too many edges I don't know I guess no because they they were only using certain types of potatoes to make sure that they weren't irregular now now they are freed from those constraints which was that's always part of the Pringle thing it's like you can like Pringles or not but they're very you know smooth and uh very thin and yeah they very consistent i mean they'll break they i get you know that's the thing is they might break but when they haven't broken they're all pretty much identical it's pretty impressive a broken wrinkle canister that's that's a nightmare but uh yeah so um all hail uh all the potatoes who haven't been part of the pringle pringle universe until now.

25:09Sounds like we're getting there. You know, starting in Poland and Belgium's on the way. No chip shortage here. No. Well, if you want product reviews from people who really use the gear that they're reviewing, then your show is live with it. Each week, we talk about the tech in our lives. It's me, it's Roger Chang, it's everybody from the DTNS family and friends, robots, laptops, mini splits. We're covering all of it. This week, our buddy, Bodie Grin, is walking us through the good and bad of the Vocaster 2. You might say, didn't you already do a Vocaster? We did, but not the Vocaster 2. So catch Live With It wherever you get your podcasts and consider directly supporting us at patreon.com slash livewithit.

26:01This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade-in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. Hi, this is Debbie Millman from the Design Matters podcast, and today's episode is sponsored by Lexus.

26:38Design isn't just about function. It is about feeling. At Lexus, every detail is intentionally designed. Master artisans don't just focus on what you can see. They obsess over what you can feel. The way the door closes. The precise click of a switch. The acoustic feedback of every control. Because a car that doesn't make you feel something is a car that stops short of amazing. Experience amazing at your Lexus dealer.

27:13All right, time for the quick headlines. You know what these do. They make you smarter. Google announced that Google Assistant will be discontinued starting September 4th on Android phones, Wear OS watches, headphones, and Android Auto when it's projected from a phone. Cars with Google built in will continue to have Google Assistant for the time being. Not all Google Assistants are gone, but most of them will be. The U.S. Voluntary Framework for Safety Testing of Frontier Models will not apply to open source models, at least according to Axios, but only to state-of-the-art closed source models that pose a national security risk.

27:56Subject models will undergo 30-day safety testing before being widely released. So there's probably going to be a delay on these frontier models. You may or may not notice, but the companies will. However, you're not going to get a look at the framework because it is considered confidential and will not be released publicly. Back to chips, but not potato chips this time. Anthropic, I know, I'm very hungry. Anthropic says it will co-design chip hardware to improve model performance. The information sources say Anthropic is considering Samsung as a partner. Anthropic going to make its own chips.

28:30Everybody's doing it. uh bgr.com notes that amazon's top selling mesh wi-fi system is tp links deco x55 ax 3000 beating out amazon's own eero as well as nest and netgear systems deco x55 sells for 190 bucks for a pack of three it's a pretty good price considering yeah everything mastercraft announced it will offer Apple CarPlay and Android Auto on Newcrest and Belize pontoon boat models starting next month. I know there's some boat people, not boat people, but people who like boats in the audience. Why don't we call it BoatPlay? BoatPlay. Yes. I mean, it's right there. Put a little spin on it.

29:18Yeah. Make it fun. You could already do CarPlay and Android Auto on boats through third-party decks. And you can do that with the MasterCraft models as well, but this is just making it easier. Yeah. Yeah. Open AI's Atlas browser is going to stop working on August 9th. And when they say stop working, like you won't be able to open it, I think, from what they're saying. So OpenAI recommends if you've got the Atlas browser, export your bookmarks, export your cookies, anything else you want to get out of there before August 9th. It's confusing because Atlas is now part of the new ChatGPT, chat gpt also the codex you know all in one thing so it's it's not gone for good but yeah but if you have the atlas app your data and you haven't moved it out you need to move it out into the atlas part of chat gpt's app yeah tell chat gpt to get your data from yeah and pull it back in yeah and just make sure they don't leave the sandbox research published in the crisper journal on Wednesday describes a technique for editing genes in a dog's skin cells to prevent it from making proteins that cause dog allergies.

30:28Geneticist Matt Walker and his team at Kindred Companion Sciences successfully performed the procedure on a pair of beagle puppies named Bailey and Elfie. Yeah. And they did it before they, when they were like in, in utero, not, not after they were born to kind of make sure that that gene wasn't there. And Matt Walker has the allergy and adopted Bailey. I don't know what happened to Alfie. I want to know who got Alfie. I'm sure somebody got Alfie. Somebody very cool. Yeah. Let us know. London's Transport Authority has issued permits for Uber and Wave, that's W-A-Y-V-E, to test their autonomous taxis in London using human safety drivers.

31:08And if you're a rider in London, you could start trying to get a ride in a wave this summer. So beating Waymo to the road. Samsung and SK Hynix are testing using chip making equipment from China's advanced micro fabric fabrication equipment for their mainland China factories amid uncertainty about obtaining other etching machines due to U.S. import restrictions. Another example of import restrictions encouraging China to get better at making the things that have been restricted. They were importing otherwise. Yeah. Version 5.05 of Google Health for iOS is out, and it lets you sync your tracked info with Apple Health on iOS.

31:52NASA scientists say that the data from the Polarometer to unify the Corona and Heliosphere, aka Punch, mission has exceeded expectations to improve the ability to predict solar storms. This helps protect satellites and avoid radio and power blackouts, among other things. Yeah, kudos to Engadget for doing a follow-up story on this. A lot of times you just talk about the launches and then nobody ever reports on it again. So this is a great example of like, hey, it's working even better than expected. And then at the Future of Memory and Storage show in Santa Clara, California, Samsung unveiled the first concept models for ZHBM.

32:31That is a high bandwidth memory design, which vertically stacks the high bandwidth memory above AI accelerators. That allegedly boosts performance up to eight times, at least according to Samsung. And then the wafer tech allows for 10 times the memory density, and this is the key to this being a story, increasing energy efficiency three times. So that reduces the power that a data center would use if they're using this memory. Well, we like to end every episode of DTNS with some shared perspectives. And today Mohan is offering some thanks. Yeah, Mohan listened to our interview with Dr. Nicky yesterday about 3D printing ear bones and said the segment on 3D printed ear was just amazing.

33:17Thank you, Dr. Nicky and Tom for making me that much smarter. RWDash also was like that that segment rocked. People seem to really like hearing about 3D printed ear bones. I mean, who wouldn't? It's pretty cool. I know. I want to hear about it. Yeah, exactly. If you're thinking about 3D printed ear bones or anything else, you have an insight into a story, some feedback for us, do share it with us. Feedback at DailyTechNewsShow.com. Big thanks to Mohan for contributing to today's show. And thank you out there listening for being along for Daily Tech News Show yourself. You can keep us in business if you're not already a patron like Mizzoula.

33:55Become one at Patreon.com slash DTNS.

34:02The DTNS family of podcasts. Helping each other understand. Diamond Club hopes you have enjoyed this program.

34:14This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade-in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. Where's your playlist taking you? Down the highway? To the mountains? Or just into daydream mode while you're stuck in traffic?

34:53with over 4 ,000 hotels worldwide Best Western is there to help you make the most of your getaway wherever that is because the only thing better than a great playlist is a great trip life's the trip make the most of it at Best Western book direct and save at bestwestern.com